China and Outsourcing*

Transcription

1 John Whalley University of Western Ontario and National Bureau of Economic Research 5 China and Outsourcing* ABSTRACT T HIS PAPER DISCUSSES the links between outsourcing and China s historic transformation involving sustained high growth and deep integration into the world economy. Recent theoretical literature tends to stress outsourcing as a switching of suppliers of intermediate inputs from domestic to foreign sources. But the reality in the Chinese case is that the integration of China and the world economy is occurring through a variety of channels of interconnection, and these channels go beyond outsourcing of intermediate products. Focusing on only one of these channels may not give a comprehensive picture. As well, there are debates in the literature about how to define outsourcing as well as attempts to measure the extent of outsourcing in a narrower sense. These debates are somewhat in contrast to public concern with the term in the (Organisation for Economic Co-operation and Development) OECD countries over what seem to be larger potential adjustments in the future in the OECD if both China s high growth and process of global integration continue. I suggest that the principal driver for these adjustments is the large pool of low-wage labour in China, restrained by immigration restrictions in the OECD, and the opportunities this creates to arbitrage wage differentials through trade in goods. The opportunities involve not only switching from local to Chinese suppliers, emphasized in current outsourcing literature, but also a wider range of transactions. These transactions involve inward foreign direct investment (FDI) flows to China to support processing trade, Chinese outward FDI (insourcing), trade occurring through middlemen An earlier draft of this paper was prepared for an Industry Canada conference held in October 2006 and organized by Daniel Trefler. 5-1

2 WHALLEY (through-sourcing), and round-trip FDI investments that originate in China and take advantage of tax and other preferences in China for inward FDI (roundabout sourcing). A wide variety of transaction channels involving Chinese entities thus impact OECD product and labour markets, and more is involved than existing outsourcing literature seems to discusses. Ultimately, labelling and categorizing the channels through which these adjustment pressures may operate may be less important than comprehending the size and scope of potential adjustments yet to come. If Chinese growth at current rates persists for several decades, we may be only at the beginning of large global adjustments. Outsourcing, as a catalyst for public concern in the OECD, has much broader connotations than the somewhat narrower concept of re-sourcing of supply. Finally, I note that China s integration into the global economy also implies adjustments outside of the OECD such as lowered FDI into Latin America and elsewhere, plus trade from elsewhere both in Asia and non-asia. These induced effects will dampen the eventual adjustment attributable to China. INTRODUCTION C HINA PLAYS A KEY ROLE for a number of reasons in the growing debate on outsourcing that is taking place in both research literature and in the public domain in the Organisation for Economic Co-operation and Development (OECD). China is the largest population economy outside the OECD. She has a large pool of low-wage and highly efficient labour that cannot move to the OECD due to immigration restrictions. Her exports in the last few years have grown at 35 to 40 percent in value terms. China s trade is now approaching 10 percent of world trade and the country is now a major presence in the global economy. China is by far the largest recipient of foreign direct investment (FDI) flows from OECD countries to non-oecd countries (around 40 percent of OECD outward FDI in 2003/2004). Low wage rates in China are widely believed to be the main driver of both these trade and FDI flows. China s trade with the OECD is approximately four times that of India, and FDI inflows to China are around seven times those entering India. Therefore, both the size and speed of change of global adjustments implied by China s growth de facto strongly suggest to OECD ears that any adjustment problems linked to outsourcing involve mainly China. This paper argues two points that link outsourcing to China s integration into the global economy. The first is that the term outsourcing as used in recent research literature appears narrower than the term appearing in public debate in the OECD. Research 5-2

3 CHINA AND OUTSOURCING literature tends to focus on outsourcing as a re-sourcing of supplies of intermediate inputs from domestic to foreign suppliers. This form of outsourcing is involved with China s integration into the global economy. However, more of the integration seems to involve marketing and distribution networks that directly arbitrage wage-rate differentials across countries through goods trade. This is more intermediation activity than outsourcing as theoretically conceived. There are also more and varied channels of interconnection through which the process of integration between China and the world economy is occurring than simply the outsourcing of intermediate products. These channels involve not only the re-sourcing of supplies of components and intermediate products to China from the OECD (conventional outsourcing as discussed in research literature) and similar activity in services (definition of outsourcing in Bhagwati et al. [2004]) but also the following: inward FDI into China relocating production from the OECD to China for export (so called vertical FDI and sometimes also equated with outsourcing); relocation of production abroad to China by Chinese firms buying OECD firms (insourcing); trade-based integration involving middlemen and distribution systems often based in Hong Kong (through-sourcing); and FDI-related activity originating in China with funds returning to China to take advantage of incentive schemes for export activity (roundabout sourcing). All of these generate adjustment pressures both in terms of potential job loss and lowered OECD wage rates; these are the pressures often associated with outsourcing in popular debate. Focusing closely on just one of these channels, when a multifaceted adjustment process accompanying China s global integration is underway, may be less productive than trying to understand the dimensions of both current and prospective future adjustments implied globally by their combined effects. The second point in the argument is that, if current high growth rates in China persist for several decades (this may be a large if), the quantitative dimensions of future potential adjustments would seem to be very large. The issue thus may be less a matter of current adjustments (from outsourcing and other re-sourcing) and more the expectation of future adjustments. I suggest that the main driver for those adjustments from China both is, and will continue to be for some time to come, the large pool of low-wage labour in China that has been made internationally immobile by OECD immigration restrictions. 5-3

4 WHALLEY Trade flows in goods that embody this low-wage labour should therefore be seen as indirect arbitrage of OECD and Chinese wage differentials. For now, foreign invested enterprises in China account for nearly half of China s exports, but only 3 percent of employment (Whalley and Xin 2006); this suggests the global integration process in China is still only at a very early stage. In turn, the trade patterns of OECD countries are already affected substantially by China s growth. Canada has seen a 60 percent increase in imports from China in five years and a growing bilateral trade deficit, but these increased imports to some extent have also displaced imports of manufactures from the United States. Canada thus has a growing trade deficit with China but a growing trade surplus with the United States, a situation that is spawning protectionist pressures in the United States against Canada. I offer some speculation as to what could be the eventual global dimensions of adjustment to the continued growth in China and attempt to relate these to the outsourcing debate. What is happening in China also cannot be divorced from what is happening elsewhere in non-oecd countries. FDI inflows to Brazil, for instance, fell from $30 billion to $10 billion between 2000 and 2003, and Brazil s share of U.S. imports fell sharply. China s growth is also a major source of concern for export performance in ASEAN countries, Mexico and other countries. Therefore, accompanying large future global adjustments from continued growth in China adjustments that are equated in public debate with outsourcing concerns will be reallocation of resources and trade involving other low-wage supplying countries. This will tend to dampen the eventual adjustment pressures from continued Chinese growth. Hence, the global adjustments involving China that may be thought to be linked to outsourcing should also be seen as part of a global process of adjustment involving all the low-wage and immigration-constrained countries and the OECD. CHINA S GROWTH AND INTEGRATION INTO THE GLOBAL ECONOMY S INCE THE MID-1980S, Chinese GDP growth performance has averaged around 9.3 percent a year according to official data. Despite continued arguments by Western economists about why this growth could not be sustained, it has prevailed and in the last two years has accelerated further. Growth rates of 35 percent for exports were recorded in 2003 and 2004, with 40 percent in Export values increased by 238 percent between 2000 and 2004 (National Bureau of Statistics of China 2005, Table 18-4). 5-4

5 CHINA AND OUTSOURCING This remarkable trade performance has been accompanied by a sharp growth in foreign direct investment (FDI) inflows into China (Whalley and Xin 2006). These inflows are discussed in more detail later in the paper but grew from around $2 billion a year in the 1990s to over $20 billion in Cumulative FDI into China now exceeds $600 billion. The resulting foreign invested enterprises (FIEs), which embody this FDI, have in turn been a major contributor to the strong Chinese export growth with export growth rates of over 50 percent in some years. The share of FIEs in Chinese exports has grown from around 15 percent in the early 1990s to around 60 percent in FIEs at the same time accounted for nearly 60 percent of China s imports in 2005, underlining the large role that these enterprises play in China s processing trade, adding value to imported products and reexporting them. Imports also represent the substantial majority of costs of these exports. These enterprises currently employ only 3 percent of the workforce but have sharply higher labour productivity. A further feature has been the large Chinese trade surpluses and, with them (and inward FDI flows), growing and large foreign reserves. China s bilateral trade surplus with the United States was $114 billion in 2005, and reserves now exceed $1 trillion. But as Gong and Li (2006) point out, China s total surplus of $101 billion in 2005 was smaller than the bilateral surplus China had with the United States. China had a trade deficit with Japan and Korea (jointly) of $58 billion, one half of the U.S. surplus, and China s imports from Japan and Korea grew sharply. Also, the majority of China s FDI enters from Hong Kong, Taiwan, Korea and Japan and the imbalances seem less severe when a regional perspective is taken of the combined China Japan Korea Taiwan Hong Kong trade with North America and Europe. It is difficult to say with any certainty how linked these rapidly changing trade patterns in China are to outsourcing. As will be seen later, both defining and measuring outsourcing is difficult. Trade data merely reflect customs clearance, and documents presented at borders provide no indication of the details behind the transactions involved. Whether suppliers have been changed for a component or whether this is wholly new order remains unknown. Some loose indication of potential links between outsourcing and China s growth in the trade can be gained from examining China s trade data from official sources. In Table 1 are China s exports for 2000 and 2004 by product category, along with percentage rates of change. 5-5

6 WHALLEY TABLE 1 CHINA S EXPORTS IN 2000 AND 2004 BY PRODUCT CATEGORY (US$ BILLION) Product Category % Change Total of which Vegetables, fruits Food, drink, tobacco Minerals, oil products Chemicals Plastics Pulp, paper Textiles and clothing Footwear Base metals Machinery and electronics Vehicles and transport equipment Precision instruments Source: China Statistical Yearbooks 2005 and 2002, tables 17-6, These data indicate there are three product categories where exports grow more rapidly than average: machinery and electronics, precision instruments, and base metals. In two of these categories (machinery and electronics, precision instruments), outsourcing might be thought likely to occur. Of those with below-average growth, vehicles and transport equipment would seem to be the most likely outsourcing category and here the growth is close to the average. Those with growth rates sharply below the average (vegetables, pulp and paper) seem the most unlikely outsourcing categories. Hence, these data might be taken as providing some support for the position that outsourcing is a factor in China s growing trade. But in Table 2, when China s trade is divided by country of export into high-wage countries from which outsourcing may be thought more 5-6

7 CHINA AND OUTSOURCING likely to originate and low-wage countries from which outsourcing is less likely, the latter dominate exports. Countries to which exports from China grew most rapidly are Russia, India, Brazil and Thailand, countries from which outsourcing seems less likely. TABLE 2 GROWTH RATES OF CHINA S EXPORTS IN 2000 AND 2004 BY COUNTRY OF DESTINATION Country of Destination % Growth of Chinese Exports between 2000 and 2004 High-wage countries United States 239 Canada 258 Germany 256 United Kingdom 237 France 267 Italy 243 Australia 257 Japan 176 Korea 246 Low-wage countries India 380 Indonesia 204 Hong Kong 226 Malaysia 315 Philippines 291 Singapore 220 Russia 407 Brazil 300 Thailand 259 Source: China Statistical Yearbooks 2005 and

8 WHALLEY These data seem to show an emerging discernible pattern: accelerating outsourcing appears to be a factor behind a change in the pattern of China s trade by commodity, but not by country. However, China s trade grows rapidly in most product categories and with all trading partners and seems largely independent of whether outsourcing is thought to be a factor or not. CHINA S TRADE AND FDI AND OUTSOURCING 1 M ORE DETAIL IS NEEDED on the links between trade and FDI patterns in order to further gauge the significance of outsourcing in China s engagement with the world economy. FDI inflows involve relocation of production. This can involve component purchase re-sourcing, as in outsourcing, or simple relocation to take advantage of low-wage costs in manufacturing. Inward FDI entering China is both vertical (with production for re-export) and horizontal (to serve the Chinese market); only the former reflects outsourcing. But this type may also be only for direct processing trade. Before 1979, FDI was prohibited in China. This restriction was lifted following China s open-door policy in 1979 when a new foreign investment law was adopted. In its early stages, FDI was restricted to China s Four Special Economic Zones and limited to equity joint ventures. Most of the FDI went into hotel construction and energy, and no outsourcing was involved. In 1984, a new foreign investment law accelerated FDI growth and a number of preferential policies were also used by both central and local governments to attract FDI. A sharp increase in FDI occurred after 1992 when China reaffirmed policies of openness and market-oriented reforms that had been introduced earlier. The resulting growth in China s inward FDI was spectacular. Annual FDI inflows in 1985 were less than US$2 billion but US$61 billion in 2004, 30 times those of 20 years earlier. Between 1985 and 1991, the annual growth rate of FDI inflows into China was 14 percent, and annual FDI inflows during this period remained less than US$4.5 billion. FDI inflows increased sharply to US$11 billion in 1992 and again to US$28 billion in 1993, with growth rates of over 150 percent in both years. By 1997, China had FDI inflows of US$49 billion. The sharpest increases in these inflows occurred in the early 1990s while there was a small decrease in FDI inflows in the late 1990s. The annual growth rate of FDI inflows increased again after China joined the World Trade Organization (WTO) in During the three years 2001, 2002 and 2003, world FDI inflows declined sharply by 41 percent, 13 percent and 12 percent respectively each year; however, China saw FDI inflow growth of 15 percent, 13 percent and 14 percent In 2004, global FDI 5-8

9 CHINA AND OUTSOURCING inflows increased only 2 percent, but China again saw an inward FDI growth rate of 13 percent. China s share of FDI flows has continued to increase in recent years and China is now the world s largest FDI recipient among developing countries and the world s second largest FDI recipient overall after the United States. By way of contrast, FDI inflows into India were only US$5 billion in A feature of China s FDI that suggests some growth in outsourcing is the increasing share that is involved with manufacturing. Before China s accession to WTO, less than 60 percent of inward FDI went to the manufacturing sector. After China joined the WTO, more inward FDI went to the manufacturing sector, reaching over 70 percent later as the share of FDI going to the real estate sector decreased. The size distribution of FIEs is a further factor in evaluating the contribution of outsourcing. By 2004, FIEs using FDI had been established, with around 50 percent of them still in operation. Among the functioning FIEs, are industrial enterprises. But the large majority are small. Only FIEs with annual sales income of over 5 million yuan (US$0.61 million) are tracked by statistical agencies in China for data purposes. The small enterprises seem less likely to be involved in component outsourcing than the larger enterprises. A large portion of China s trade is conducted via these FIEs, which might be taken to indicate outsourcing with physical presence with transactions internal to the firm. A government estimate is that 46 percent of the output of FIEs in the manufacturing sector is for export from China, while the figure is only 16 percent for non-fies. Since the early 1990s, the export growth rate of FIEs has been over 30 percent in most years, with over 40 percent in some years a much higher growth rate than that for non-fies. The ratio of exports from FIEs has increased steadily since 1990, reaching 30 percent in 1995, more than 40 percent in 1996, 50 percent by 2001, and 60 percent by Along with rapid export growth, FIEs have also accounted for progressively more of China s imports. FIEs are heavily involved in component import, assembly and re-export as processing trade rather than the type of outsourcing stressed in theoretical literature. In 2004, the ratio of imports by China s FIEs to total imports was about 60 percent, although this ratio is more volatile than that for exports. The export ratio was 55 percent in 1998, declined to 37 percent by 2001 and rebounded by almost 20 percentage points in Since 1990, FIEs have also accounted for most of China s export growth. Without the growth in FIE exports, China s export growth rate would have been only about 10 percent in most years after 1990, and negative in some years. Whalley and Xin (2006) note the sharp change in the share of 5-9

10 WHALLEY China s exports of FIEs from less than 2 percent in 1990 to nearly 60 percent in Lemoine (2001) reports that most of the vertical- and exportoriented FDI inflows to China originate from elsewhere in Asia (primarily from Hong Kong, Korea and Taiwan) where outsourcing would seem to be less of a major issue. Many of these businesses are owner-managed from initial investments, with the core business growth occurring inside rather than outside China. FDI in China from North America and Europe is claimed to be more heavily horizontal and aimed more at the Chinese domestic market than exports, indicating further qualification to the view that outsourcing is behind FDI growth. Fung (2004) reports an estimate that U.S. origin FIEs sold more than 80 percent of their products locally in 2002, while FIEs of Japanese origin sold only 45 percent of their production locally. And while Hong Kong, Macau and Taiwan accounted for 66 percent of FDI inflows between 1979 and 1992; although this share declined later, it still was around 40 percent in Finally, because labour productivity in China s FIEs is sharply higher than elsewhere in the economy maybe by a factor of four compared with the rest of the industrial sector (Whalley and Xin 2006) one can argue that the primary attraction of FDI is access to a high-quality, low-wage labour pool that they can use effectively with foreign technology in processing trade of various forms. Fung (2004) notes that in 2003 foreign firms accounted for 56.2 percent of China s imports and 54.8 percent of its exports. In 2004, FIEs accounted for only 3 percent of China s workforce but 22 percent of GDP and over 55 percent of exports. These figures do not point to large FDI inflows merely reflecting outsourcing, but rather reflecting direct access to the low-wage labour in China for cost savings in production and processing trade. A further feature that weakens the argument for outsourcing as a dominant driver of growing exports is the large amount of China s trade conducted through middlemen and distribution arrangements centred in Hong Kong, Korea and Taiwan. Fung (2004) notes that 28.3 percent of China s exports were re-exported via Hong Kong and 21.9 percent of Chinese imports were imported via Hong Kong. Trade through middlemen would seemingly be unlikely to reflect outsourcing of the type stressed in theoretical literature, although to some degree this is a reflection of trade in textile and apparel and the impact of bilateral multilateral fibre agreement (MFA) quota restraints on China in the U.S. and EU markets. As Whalley (2006) notes, the termination of the MFA in January 2005 has seen large increases in China s exports of clothing to the United States and European Union, accompanied by sharp falls in exports to Hong Kong as these quotas were removed. But 5-10

11 CHINA AND OUTSOURCING at the same time, this Hong Kong trade also involves considerably more products than just apparel especially electronics. A considerable amount of China s trade is related to processing and assembly, with the common organizational form being sourcing via retailers in OECD and other markets abroad. Typically, processed exports largely embody labour as the domestic value added and have smaller value-added content than non-processed exports. Electronics account for a significant fraction of this processing activity, along with clothing. Processed exports are also typically products for final rather than intermediate sale, again typified by consumer electronics and clothing. Fung (2004) reports official estimates that 55.2 percent of China s exports in 2003 were processed exports, while 39.5 percent of the country s imports were inputs for processing. Feenstra and Hanson (2005) report similar estimates, that 55.6 percent of China s exports between 1997 and 2002 were processed. However, Fung (2004) reports that, in contrast to the 1990s, the majority of foreign investment inflows to China now take the form of wholly foreign-owned entities rather than joint ventures. This suggests that, if outsourcing is involved, it is internal to the firm and not of the type emphasized in theoretical literature. Fung reports that 64 percent of FDI in 2002 was of this form. On the other hand, and as Fung (2004) notes, there is substantial growth in China s technology trade. He notes that 2003 data from China s Custom Statistics report 25.9 percent of China s exports as high technology products, along with 28.9 percent of imports. He suggests that two-way trade of high technology products would now seem to be substantial and that processing-related activities in this area would seem to be linked more directly to outsourcing. The geographical concentration within China of both China s trade and FDI inflows is a further element related to outsourcing. Fung (2004) notes that Guangdong in 2003 accounted for 31.7 percent of China s imports and 34.9 percent of China s exports. The majority of FDI inflows still go to the eastern and southeastern coastal areas, although the picture is changing. If foreign outsourcers find comfort in the close geographical proximity of fellow travellers or find advantage in hiring away newly trained low-wage labour, this tends to elevate outsourcing as a source of trade growth. While direct studies of outsourcing in China are few, some do exist for Japanese firms operating in China. While Tomiura (2005) finds that relatively few Japanese firms outsource to China, Xing (2004) found that Japanese-affiliated manufacturers in China operating as subsidiaries sold about one third of their products back to Japan. In industrial machinery, more than half of the products were exported to the Japanese market either as final goods or intermediate inputs. In the textile and transportation equipment sectors, ratios of re-export 5-11

12 WHALLEY increased between 1997 and In 1997, only 31 percent of textile products were produced for the Japanese market while by 2002 the share had risen to 47 percent. In transportation equipment, Japaneseaffiliated manufacturers also exported 47 percent of their products to Japan in 2002, much higher than the 27 percent re-export ratio in Re-exports increased by 23 percent in 2000 and continued to grow in 2001 and 2002 with 36 percent and 16 percent growth rate respectively. This suggests growing Japanese-affiliate production in China for the Japanese market, but it is unclear whether this is outsourcing or merely processing to utilize lower-wage labour. All these studies and data thus suggest that China has indeed become a significant production base for both final products consumed abroad (largely in OECD markets) and intermediate inputs purchased by foreign firms. Rauch (2003) characterizes China as now being part of a global supply chain network in certain key industries and engaged in both importing and exporting of various components and parts. This suggests increasing vertical integration between OECD and Chinese industrial activity, and the creation of linkage between firms in the two economies. A significant portion of this is affiliate activity and a further portion involves owner-operated firms from Taiwan and Hong Kong. It is unknown how much of this reflects relationship-specific production chains as stressed in theoretical outsourcing literature. It clearly seems involved. However, the issue is whether the majority merely involves linkage of low-wage production to global networks to take advantage of low-wage costs for processing trade. CHINESE OUTWARD FDI AND CHINESE INSOURCING 2 C OMPARED WITH INWARD FDI, Chinese outward foreign direct investment (FDI) has been relatively small until recently, in the US$3 billion to US$4 billion range in It has been heavily concentrated on both greenfield and joint venture activity abroad, much of it focused on Hong Kong. As such, it seemingly has little relation to outsourcing. But, in the last year, direct acquisition of foreign firms by Chinese entities involving potentially large transactions have become prominent, in some instances in the US$15 billion to US$20 billion range, and with a focus well beyond Hong Kong. Examples include the Lenovo buyout of IBM's PC business, CNOOC's bid for Unocal, a prospective bid by MinMetals for Noranda, the Haier Group bid for Maytag, and others. Rather than conventional outsourcing, these transactions represent insourcing or the relocation to China of foreign corporate activity through Chinese acquisition. In part, the approach is for the Chinese acquirer to combine the acquired foreign distribution system with growing lower-wage Chinese manufacturing output that 5-12

13 CHINA AND OUTSOURCING slowly displaces higher-wage supplies abroad. Access to foreign brands in the distribution process is an element behind this. At first sight, an upsurge of new outward Chinese FDI may strike outside observers as strange. China is, after all, still a relatively capitalscarce economy with a large pool of low-wage labour. As well, development policy in China remains focused on attracting inward platform FDI to combine with low-wage labour to fuel further export and GDP growth. To begin exporting large amounts of capital through large foreign acquisitions when this broad stance of development policy toward inward FDI remains unchanged calls for an explanation. Chinese FDI has previously been concentrated outside of the OECD. In 2003, 80 percent was in Asia, 14.3 percent in Latin America, 1.7 percent in North America, 1.5 percent each in Europe and Africa, and 1.4 percent in Oceania. The top five destinations for Chinese investment in 2003 were Hong Kong, Cayman Islands, Virgin Islands, United States and Macao (in that order), followed by Australia, Korea and Singapore. It has also been concentrated on three sectors: information technology, computers and software (33 percent in 2003); distribution, wholesale and retail (20 percent); and mining (18 percent). In 2004, Chinese outward investment was only US$3.6 billion, only 5 percent of the FDI flowing into China, despite the growth rate of outward FDI being twice that of inward FDI for several years prior (in 2004, 27 percent and 13 percent respectively, year on year). A recent joint study undertaken by the Chinese Academy of International Trade and Economic Cooperation (CAITEC) and the Welsh Development Agency (WDA) used a questionnaire to investigate key factors involved in Chinese outward FDI. The study concluded that a range of factors were involved, including market expansion, implementation of long-term development strategies for firms, more access to technology, accessing advanced management methods, avoiding trade barriers, taking advantage of foreign preferential investment policies, achieving cost reductions, acquisition of material inputs (resources), and transferring excess production capacity abroad. The use of overseas investment as a way around official Chinese restrictions on access to foreign exchange and foreign capital markets was also cited as a factor. Other studies have pointed to non-economic reasons for Chinese enterprises investment abroad, such as the possibility of gaining residency rights and other benefits in the host country for managerial staff (such as health services, social security and access to education). But the decision by a Chinese company to buy a foreign company is also often linked to an intent to move manufacturing activity to China to benefit from lower labour costs while keeping existing distribution networks of the acquired business in the 5-13

14 WHALLEY host country. These considerations are similar to outsourcing, involving foreign firms in their activities in China. Recent widely publicized Chinese bids for large firms in the OECD are also portrayed in Chinese media as involving a prestige factor: going abroad either to buy a recognizable foreign brand (Lenovo's takeover of IBM) or to build their own brand's identity by establishing manufacturing plants in the target country (such as Haier's factory in South Carolina). However, recent Chinese outward FDI also reflects the official Chinese government policy to encourage domestic enterprises to invest abroad. Government regulatory approval processes for overseas investment projects have been simplified significantly in recent years, increasing outward FDI flows. Chinese companies are also supported by low-interest loans if their overseas activity involves, among others, resource exploration and acquisition of advanced foreign technology. This use of low-interest loans made available to Chinese state-owned enterprises (SOEs) is supported by China's large and growing foreign reserves. These reserves today stand at close to US$1 trillion, reflecting both large Chinese trade surpluses in recent years and inward foreign investment. With concerns in China over the security of supply of resource inputs as well as the prospect of the U.S. dollar possibly falling further (and its impact on China s large holdings of U.S. Treasury bills in the reserve portfolio), deploying Chinese reserves in this way is seen in China as a reasonable policy. Chinese government policy is now to use outward-oriented investments, mainly by large SOEs, to secure access to resources and raw materials (especially iron ore, coal, oil and natural gas); acquire new technology for transfer back to China; expand Chinese export markets; and strengthen international relationships with, and gain more influence in, other countries. In 2003, SOEs accounted for only 43 percent of total Chinese investments abroad with limited liability, shareholding and private companies together accounting for another 43 percent. With the recent large-scale acquisition activity abroad, these proportions seems poised to change. Political resistance to these transactions has emerged in the OECD with questions of subsidies, lack of transparency, and national security all being raised. National security issues regarding foreign acquisitions are not new; in the United States, they go back to the Exon-Florio provisions of the American Defense Production Act of 1998 following concerns in the United States in the late 1980s over Japanese buyouts. However, issues of subsidization of foreign acquisitions through lowinterest loans from central banks and the transparency of organizational form of acquiring entities (SOEs) are new. 5-14

15 CHINA AND OUTSOURCING Insourcing as well as the more conventional outsourcing thus seems poised to become part of the growing linkage between China and the global economy. For now, outward FDI from China (insourcing to China) is relatively small, but the size of China s foreign reserves suggests this could change significantly in future years. Here, the linkage to distribution and marketing networks seems even more marked than in the inward FDI area. LITERATURE ON OUTSOURCING AND OUTSOURCING TO CHINA T HUS FAR I have repeatedly used, but not defined, the term outsourcing. This term has been the focus of both theoretical literature spawning models of a firm s decision to outsource and the impact on industry equilibrium outcomes and empirical literature that seeks to measure it. A sharp contrast exists between the discussion of outsourcing in the literature and the term s use in more popular debate in the OECD on China and outsourcing. Broadly defined, the term outsourcing in research literature is taken to refer to the purchase of goods and services that were previously produced inside a company. The company providing the goods and services can be located in the same country (domestic outsourcing) or outside the country where the purchasing company is located (international outsourcing). The international component of outsourcing relates to switching sourcing from within to between countries, and China s size and low-wage structure make that country a major element in this literature discussion. Feenstra and Hanson (2005), for instance, explicitly formalize and estimate a model of ownership and control in analyzing outsourcing to China. There are many variants of and elaborations on constructions of outsourcing in the literature. Grossman and Helpman (2002) emphasize the vertical disintegration of production that seems to be involved; however, they stress that, to them, outsourcing means more than the purchase of raw materials and standardized intermediate products. It also means finding a partner with which a firm can establish a bilateral relationship and having the partner undertake relationship specific investments so that it becomes able to produce goods and services that fit the firm s particular needs. They then model the firm s choice of relationship-specific partners in either a technologically and legally advanced North or a low-wage South, emphasizing the search process involved, the need to convince potential suppliers to customize products for their needs and the incompleteness of contracts available for enforcement. 5-15

16 WHALLEY Bhagwati et al. (2004), in contrast, focus their discussion of outsourcing on services, specifically long-distance purchase of services by electronic media such as telephone, fax or Internet, and discuss alternative cases where providers and purchasers of services have differing degrees of mobility along the lines of Modes 1 to 4 of the General Agreement in Trade in Services (GATS). The motivation they offer is the U.S. president s Council of Economic Advisers in 2004 characterizing the outsourcing of professional services as a new type of trade. This discussion also reflects a perceived rapid growth of outsourcing in the information technology area to India. The OECD (2006), in a recent survey of outsourcing literature, also points out that outsourcing and offshoring are terms which are frequently used interchangeably. Offshoring refers to the purchase of goods and services, previously produced inside the purchasing company, from companies in locations outside the country. To the OECD, the terms include not only international outsourcing but international insourcing with the foreign affiliates of domestic parent companies exporting to their parents. These and other literature-based notions of outsourcing essentially reflect attempts to both (i) delineate and categorize the various channels through which economic integration is proceeding globally between various economies, and (ii) better understand the changing nature of global production processes in both manufacturing and services. But exactly how this pattern of production is changing is conjecture, and theoretical literature often relies on anecdotes. Examples such as the global production of American cars, the globally fragmented production of Barbie dolls, and other items are frequently mentioned. But for many years, large OECD manufacturing firms have had many component suppliers and outsourcing as such is not new: Boeing is reputed to use over a million components in aircraft assembly, and General Motors has long had thousands of component suppliers. The issue is how the process of working with these component suppliers is changing for large OECD manufacturers as global integration proceeds across national boundaries involving low-wage economies. China is central to this. On the one hand, manufacturers seemingly prefer component suppliers to be geographically close so it is easier to monitor them as well as communicate with them over product quality and the frequent small changes to components that are required when meeting changes in market tastes. Just-in-time inventory management is also easier with suppliers who are geographically close, and there can be an accumulation of person-to-person links with individuals in supplying firms that makes switching suppliers costly. But if low-wage (and hence low-cost), reliable and high-quality component suppliers become available elsewhere, the cost savings can justify re-sourcing across 5-16

17 CHINA AND OUTSOURCING national borders even with the geographical barriers involved. The impression of recent outsourcing activity involving China is that the improved infrastructure in China, the growing reputation for product quality and on-time delivery, the reliability of middlemen in organizing transactions (often via Hong Kong) and the improved local legal structure are all important elements which go beyond wage-cost differentials. Public concern in the OECD over outsourcing, however, stands in some contrast to the research discussion in literature. The concern in the OECD focuses largely on potential job loss and downward wage pressures as outsourcing occurs abroad rather than on seeking to provide analytics for firms decision making on vertical cross-border unbundling (or disintegration) of production processes. As such, the policy concern is about the adjustments implied by international integration, not about understanding exactly what form the production unbundling takes. Production disintegration via intermediate product outsourcing is but a part of this concern. Given this adjustment focus, there seems to be little reason why debate on outsourcing should limit itself to re-sourcing of intermediate products and components. OECD retailers re-sourcing supplies from domestic to foreign firms generate adjustment pressures in the OECD in the same way that cross-border production unbundling does. There are also more channels through which the impact of integration processes occur than just the re-sourcing of component suppliers, the channel stressed in outsourcing literature. As discussed above, these channels include firms in low-wage countries such as China buying an OECD firm and keeping the firm s distribution system in the OECD, but moving production of manufacturing back to China (one can term this insourcing). Also, as noted above, a considerable portion of China s integration into the global economy has been facilitated by trade transactions orchestrated and conducted via middlemen in Hong Kong, Taiwan and Korea. The size of these transactions is large, as are the trade impacts. There are speculative estimates that profits from Mainland China accruing to Taiwanese companies are around 70 percent of the profits of companies quoted on the Taiwanese stock exchange. This set of activities might be called through-sourcing to distinguish it from conventional outsourcing. A further element of the global integration process in which China is involved is its measurement, and whether constructed transactions inflate the seemingly real size of the integration process. Investment funds of Chinese origin flow abroad and return to China to benefit from preferential policies toward foreign invested enterprises. Foreign invested enterprises receive a sharply lowered enterprise tax rate (one half in certain types of zones and areas), tax holidays, exemption from 5-17

18 WHALLEY municipal surtaxes and other benefits. Typically, to qualify for this considerable preferential tax treatment, threshold tests of export sales must be met. Xiao (2004) reports an estimate that the ratio of roundtrip FDI via Hong Kong to total FDI inflows to China could be 40 percent. Again, the transactions seem large and have significant impacts on the perceived linkage of China s trade to outsourcing. This could be termed roundabout sourcing, also for want of a better term. In current research literature, it is not only the definition of outsourcing that is somewhat ambiguous and seemingly narrow. In thinking about the term outsourcing as applied to China, one can come to the conclusion that the wide variety of channels of interaction with the global economy are at issue and that these channels go beyond those focused on by conventional outsourcing literature. And if the concern in OECD countries over outsourcing is about adjustment costs, rather than about understanding the changes in the nature of production processes, then all of these channels would seem to merit discussion when outsourcing is under scrutiny. A further element in the literature on outsourcing reflects the attempts to measure outsourcing and assess its impact on OECD wage rates. As China is central to such impacts, given its size, this component of literature is worth noting here. This work has proceeded parallel to the analytical literature referred to above. Direct measurement of outsourcing is difficult, and researchers typically examine detailed trade data and classify product categories involving either intermediate or final products. Growth in intermediate relative to final products is typically taken to reflect increased outsourcing. These studies usually show increased outsourcing. Abraham and Taylor (1996) document an increase in the outsourcing of business services in 13 U.S. industries using this approach. Campa and Goldberg (1997) also measure outsourcing of imported intermediate inputs in this way for various industries in Canada, Japan, the United Kingdom and the United States. These authors show that, except for Japan, there is a doubling in the share of imported inputs between 1975 and 1996 for all manufacturing in the United States and that the United Kingdom also demonstrates a large increase in outsourcing. Feenstra (1998) measures all imported components used in production by U.S. firms and computes various measures of outsourcing, also arguing all have increased since the 1970s. Given China s size of trade, the presumption would appear to be that China is involved in a considerable portion of this activity. While these findings indicate a change occurring in the volume of outsourcing as understated in conventional literature, the issue remains of how large the initial base is (around which this change is occurring). This is less well researched. Tomiura (2005) studies Japanese firms and 5-18

19 CHINA AND OUTSOURCING their outsourcing activities and reports that relatively few Japanese firms outsource across national boundaries. The impression from U.S. business activity is that still today a large part of component purchases occur from smaller component suppliers located in relatively close geographical proximity to larger scale manufacturing establishments so close coordination and contact can be maintained. A further issue in the Chinese case is whether more significant components of Chinese foreign trade involve final products than is true of other countries, even if the intermediate component trade is growing. Significant Chinese exports occur in clothing and assembled electronics. And much of Chinese export activity builds on relationships with retailers and distributors in the OECD that are anchored in Hong Kong, Taiwanese, and Korean businesspeople with long experience in how these distribution systems work. There is speculation, for instance, that Wal-Mart is the eighth largest trading partner of China and may have more trade with China than the whole of the United Kingdom. So while there are no firm data to confirm it, the possibility exists that outsourcing as conceived of in the theoretical literature may for now still be relatively small even if it is growing quite rapidly. Finally, there is also discussion in the literature on the distributional impacts of outsourcing. There are two distinct parts to the literature: one is on the indirect effects on bargaining power of unions in OECD labour markets due to the threat of outsourcing production; the other is on the direct effects on OECD wage rates as production increases offshore. The first of these effects is emphasized in Rodrik (1999) and Gaston (2002). Results reported in recent empirical work by Dreher and Gaston (2005) indicate that various measures of globalization are negatively correlated with union bargaining power and union membership. In terms of outsourcing s impact on wage rates, Feenstra and Hanson (1996; 1997) find that outsourcing has led to an increase in the wage of skilled versus unskilled labour in both the United States and Mexico. Feenstra and Hanson (1999), in a subsequent paper, find that U.S. outsourcing has raised the real wage of non-production workers by 0.16 percent a year and of production workers slightly (by 0.01 percent a year). Outsourcing involving the United States and Mexico thus appears to be a win-win situation. For now, wage impacts in the OECD according to this literature thus seem mild and positive, reflecting the efficiency gains from greater outsourcing raising the real wage of both skilled and unskilled labour. None of this final strand of literature relates directly to China and so the distributional implications of outsourcing involving China remain unclear. One central difficulty is the inability to attribute distributional consequences, say, in U.S. labour markets, to bilateral 5-19

20 WHALLEY outsourcing involving China if outsourcing is occurring simultaneously to several countries and there is an integrated U.S. labour market. THE OECD ADJUSTMENT IMPACTS OF OUTSOURCING, AND THE POTENTIAL FOR THEM TO GROW IN THE DECADES AHEAD A S I NOTE ABOVE, public concern in the OECD over outsourcing is not only a reflection of current job losses that might be attributed to outsourcing and production relocation through outward FDI. It also (and seemingly predominantly) reflects expectations about how large the adjustments attributable to these integration processes might be in the future. China figures for now as the central element in these concerns since China is so large and her gross domestic product (GDP) and trade are growing more rapidly than in any other low-wage economy. The issue is whether we are at an early stage of a historic transformation in which large portions of global manufacturing and service activity progressively relocate to low-wage economies to arbitrage wage differences supported by OECD immigration restrictions; or whether this is only a more limited adjustment that will slowly dissipate in future years. Robert Mundell, in a well-known paper some decades ago (1957), showed a formal equivalence between goods flows in the presence of fixed factors across countries that achieved factor price equalization, and freely flowing factors across national borders where goods are immobile. He showed that these two achieve the same result. To some degree, both global integration and current developments involving China mirror this situation. With international restrictions on labour flows, flows of goods between countries are generated as a way of indirectly arbitraging international wage differentials through goods that embody labour. Were the world economy instead to be a single integrated economy in which factors, including labour, could flow without restraints across national borders, international factor price differentials (and especially wage-rate differentials) would greatly narrow and trade in goods would fall significantly. China is the largest of the low-wage economies currently outside of the OECD. The potential for even larger adjustments occurring globally would seem to be large as the Chinese integration into the world economy, and the world s integration into China, continues and even accelerates. The amount of labour that has been involved thus far in the trade generated by these low-wage differentials is still small since FIEs in China, while accounting for nearly 60 percent of China s exports, 5-20